[cherry pick 2.0-beta] some bug fix commit from develop to 2.0-beta (#162)

* fix distributed bug (#160)

* update docs configs (#161)

README.md

@@ -45,6 +45,8 @@ GAN-Generative Adversarial Network, was praised by "the Father of Convolutional
 * [Photo2Cartoon](./docs/en_US/tutorials/photo2cartoon.md)
 * [Wav2Lip](./docs/en_US/tutorials/wav2lip.md)
 * [Super_Resolution](./docs/en_US/tutorials/super_resolution.md)
+* [StyleGAN2](./docs/en_US/tutorials/styleganv2.md)
+
 
 ## Composite Application
 

README_cn.md

@@ -46,6 +46,8 @@ GAN--生成对抗网络，被“卷积网络之父”**Yann LeCun（杨立昆）
 * [Photo2Cartoon](docs/zh_CN/tutorials/photo2cartoon.md)
 * [Wav2Lip](docs/zh_CN/tutorials/wav2lip.md)
 * [Super_Resolution](./docs/en_US/tutorials/super_resolution.md)
+* [StyleGAN2](./docs/en_US/tutorials/styleganv2.md)
+
 
 ## 复合应用
 

configs/esrgan_psnr_x4_div2k.yaml

@@ -18,8 +18,8 @@ model:
 dataset:
   train:
     name: SRDataset
-    gt_folder: data/DIV2K/DIV2K_train_HR
-    lq_folder: data/DIV2K/DIV2K_train_LR_bicubic/X4
+    gt_folder: data/DIV2K/DIV2K_train_HR_sub
+    lq_folder: data/DIV2K/DIV2K_train_LR_bicubic/X4_sub
     num_workers: 4
     batch_size: 16
     scale: 4

configs/pix2pix_cityscapes_2gpus.yaml

@@ -64,7 +64,10 @@ dataset:
     preprocess:
       - name: LoadImageFromFile
         key: pair
-    - name: Transforms
+      - name: SplitPairedImage
+        key: pair
+        paired_keys: [A, B]
+      - name: Transforms
         input_keys: [A, B]
         pipeline:
           - name: Resize

docs/en_US/tutorials/super_resolution.md

@@ -51,8 +51,26 @@
   ```
     python data/process_div2k_data.py --data-root data/DIV2K
   ```
+  When the program is finished, check whether there are ``DIV2K_train_HR_sub``, ``X2_sub``, ``X3_sub`` and ``X4_sub`` directories in the DIV2K directory
+  ```
+    PaddleGAN
+      ├── data
+          ├── DIV2K
+                ├── DIV2K_train_HR
+                ├── DIV2K_train_HR_sub
+                ├── DIV2K_train_LR_bicubic
+                |    ├──X2
+                |    ├──X2_sub
+                |    ├──X3
+                |    ├──X3_sub
+                |    ├──sX4
+                |    └──X4_sub
+                ├── DIV2K_valid_HR
+                ├── DIV2K_valid_LR_bicubic
+              ...
+  ```
 
-  Prepare dataset for realsr df2k model:
+#### Prepare dataset for realsr df2k model
   Download dataset from [NTIRE 2020 RWSR](https://competitions.codalab.org/competitions/22220#participate) and unzip it to your path.
   Unzip Corrupted-tr-x.zip and Corrupted-tr-y.zip to ``PaddleGAN/data/ntire20`` directory.
 

ppgan/engine/trainer.py

@@ -144,9 +144,9 @@ class Trainer:
         self.best_metric = {}
 
     def distributed_data_parallel(self):
-        strategy = paddle.distributed.prepare_context()
+        paddle.distributed.init_parallel_env()
         for net_name, net in self.model.nets.items():
-            self.model.nets[net_name] = paddle.DataParallel(net, strategy)
+            self.model.nets[net_name] = paddle.DataParallel(net)
 
     def learning_rate_scheduler_step(self):
         if isinstance(self.model.lr_scheduler, dict):